- From: Roy T. Fielding <fielding@liege.ICS.UCI.EDU>
- Date: Wed, 05 Jun 1996 23:11:25 -0700
- To: Daniel DuBois <dan@spyglass.com>
- Cc: jg@w3.org, http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
> NO! I am not going to code our server so that it searches through the last
> modified times of every single variant looking for the latest before it
> determines the result of an If-Modified-Since test.
Oh, give me a fricken break -- I already showed how it can be done without
any searching.
It is a protocol error because If-Modified-Since and If-Unmodified-Since
depend on this definition in their use of Last-Modified date stamps for
the IMS and IUS values. If the value is allowed to reverse itself
(as would happen with variants having different LM times), then the
user agent will not get the right result.
If we are to CHANGE the definition of Last-Modified, then we also need
to change the definition of IMS such that it must use the value of the
received Date header field when IMS is used for cache validation.
I don't care which we go with, but we cannot change one without changing
the other.
...Roy T. Fielding
Department of Information & Computer Science (fielding@ics.uci.edu)
University of California, Irvine, CA 92717-3425 fax:+1(714)824-4056
http://www.ics.uci.edu/~fielding/
Received on Wednesday, 5 June 1996 23:54:59 UTC